Peak erodes to mound

One morning last summer I was looking at the overnight police activity (as usual), and noticed what appeared to be an unusual day, in that police dispatches had been low during the wee hours of the  morning when I would have expected to see a spike. There's always been a peak around the time bars are closing and people who have been drinking are heading out. It's not much of a peak anymore: it's eroded to a mere mound.

The smoothing of the curve I noticed in August is not an anomaly: there has been a change, and a big one. The following two charts demonstrate the change. These temporal heat charts, organized by day of week in the columns and hour of day in the rows, show the relative volume of police dispatches during the 168 one hour weekly time slots. Colors are assigned by standard deviation breaks. Dark red means a very high peak compared to the mean, deep blue means a very low hour. The top chart is for the year ending March 31, 2016. the bottom one covers calendar year 2007, the year my blog started.

The change evident from these charts is striking. The row totals reveal the significance of the change, particularly if you consider that Lincoln's population (of both people and liquor licenses) has increased significantly in the past nine years. Bar break just isn't quite what it used to be, and that's a good thing.

Problem Resolution Team app

Back in the mid-1990s, the Lincoln Police Department was diving into problem-oriented policing, which at the time was still a fairly new concept: instead of responding over and over again the the same place or problem, try to identify the underlying issues and work to resolve those. Several examples of such work can be found on my blog, by following the POP tag in the label cloud. Another great source for information is The Center for Problem Oriented policing, popcenter.org.

One of the early practitioners of problem-oriented policing in Lincoln was Capt. Jon Briggs, who commanded the Northeast Team at the time. Some of the problems his officers were working on required the assistance of other agencies, such as the Health Department, Aging Services, the Building & Safety Department, Animal Control, and the Law Department. Jon saw a real need to coordinate and collaborate across agency boundaries.

This need became a concept paper, which we presented Mayor Mike Johanns, and Lincoln's Problem Resolution Team was born--with Capt. Briggs as it's chair. In the ensuing 20 years, the PRT has become institutionalized in Lincoln, and several LPD managers have served as the chair. LPD Crime Analyst Char Estes provides the technical support, among her other duties. The team has experienced many successes in resolving chronic issues at problem properties, and today many other cities have similar inter-disciplinary teams of this type.

Longtime readers of this blog know that I am something of a minor league GIS geek. Every now and then I'll dive into a GIS project, which is an opportunity to work on something that requires an entirely different skill set than my normal job duties. This week, I spent several evenings working on a project for the PRT. With a little help from Jeff McReynolds, Lincoln's GIS program manager, I was off to the races in an effort to create a mapping application for tracking current PRT properties.

I used ArcGis Online to build a web mapping application that displays the location of these properties. A click on the icon brings up the details, including the most recent photo from the Lancaster County Assessor's Office, and a link to more detailed information about the property. I think it will be a nice tool for the PRT, and it's certainly a good example of the utility of ArcGIS online if even a GIS hobbyist like me can do it.

Many crime analysts read the Director's Desk, and most of those are involved in GIS work as part of their duties. If you have yet to explore creating web mapping applications, I suggest you do so. This is only one of many we use for a variety of purposes in public safety: CCTV cameras, parcel lookups, street finder, fire hydrant locations, fire pre-plans, P3i, stream gauges, and much more. Web mapping applications are ideal when you need a simple, single purpose application quickly.

A good starting point (other than visiting arcgis.com) would be to look around the GIS community in your own jurisdiction: the county assessor, public works department, parks & recreation, building inspections agency, and so forth. You may find that other municipal GIS technicians are already deploying web mapping applications, and can offer you some assistance in getting started.

Motown lowdown

I was in Detroit most of the week, helping teach a seminar on crime analysis for police executives, funded by the Bureau of Justice Assistance, a division of the United States Department of Justice. I agreed to serve as a so-called "subject matter expert" at two or three of the ten seminars planned around the country.

This one was bigger than Oregon, with about 20 agencies in attendance, such as Grand Rapids, Flint,  and Detroit. It was a great group, and I had some excellent offline discussions with the Bureau of Justice Assistance, the Michigan Department of Corrections, Canton Police, Grand Rapids PD, and the Detroit Public Schools.

A highlight for me, though, was on Friday morning after the conference. With late flights and time to kill, my two co-instructors and I managed to invite ourselves to the Detroit Police Department Crime Analysis Unit, on the seventh floor of the spanking new downtown police headquarters. Hey, what do you know: it's a Department of Public Safety joint police and fire facility! Sgt. Sloan and Deputy Chief Levalley were our hosts.

I was impressed with their operation, which, despite the huge size differential, employs a quite similar thought process and strategies to those we use in Lincoln to make information available to our personnel in a straightforward and accessible interface relying upon a web browser and an Intranet.

Dr. David Martin, from Wayne State University's Center for Urban Studies, spent quite a bit of time showing me how he has helped DPD create a web-based Intranet interface to the police RMS and other information systems. It is very similar in concept to the strategy that has guided Lincoln's police information resources since the mid 1990's, and cutting edge. I hope I left Detroit with at least one good idea from Lincoln: I know they left me with one.

Modus operandi

Latin for "method of operation," modus operandi is commonly considered in criminal investigations, when officers or analysts assess potential connections between similar cases. Criminals frequently have behaviors or styles that form a signature, a certain way of doing things.

Last week, I helped teach a seminar in Oregon, "Crime Analysis for Police Executives," sponsored by DOJ's Bureau of Justice Assistance and the International Association of Directors of Law Enforcement Standards and Training. The training teamed crime analysts with their command staff. As with most trainings, I learned as much as I imparted, a lot of which took place in the informal conversations during breaks.

When you rub shoulders with a group of crime analysts, they often lament the lack of solid M.O. data in database fields, which makes queries difficult, impossible, or not entirely accurate. Incomplete entries, or even ignoring the M.O. fields entirely is a rather common problem. If you want police officers to use those fields reliably, there are two things that can be done to make this more likely,

First, make sure officers understand the value and importance of the data. Do some training, so officers understand that the only way you can provide a map or list of robberies committed by a suspect who threatens victims with a knife is if the officers have completed the weapon type and usage fields in their Incident Reports. When officers see the results of providing the M.O. data, they will be encouraged to collect it with more diligence.

Second, ensure that the system for collecting M.O. data is as straightforward as possible. I have seen some multi-page M.O. forms that make my eyes glaze over. Keep the number of fields reasonable, and the selection of values for those fields concise. Ideally, an officer completing a report on an assault would not have to wade through all of the fields that only pertain to other crimes, like vandalism, auto theft, or robbery in order to get to those that apply to assault.

Several years ago, when we were developing an online Incident Report at LPD, we were inspired by TurboTax: if you have no self-employment income, you never see all the additional fields associated with that particular circumstance. Similarly, if you are reporting a burglary in Lincoln, and have entered the incident code for that crime type, the only M.O. fields you'll see are those that relate specifically to burglary: point of entry, method of entry, day/night, and victim activity. The choices in the dropdown are relatively few. It's great data when an officer or analysts wants to know something like, "Where are any other burglaries where entry was made through an overhead door left standing open?"

Here are a few screen shots clipped from the Incident Report entry screen, that show the M.O. fields and dropdown values for burglary (click image for larger view.)

 

Anniversary of a memorable case

I interrupt my current blog series on suicide data for a moment, to reminisce about one of the most memorable days in a few 40 year careers policing Lincoln. Ten years ago today, on the morning of January 14, 2005, Officer Scott Arnold and Sgt. Ken Koziol nabbed a serial bank robber culminating a rather incredible (but very short) special project. I was in my office glued to the police radio that morning, having a hard time believing what I was hearing--and then smiling from ear to ear.

This was without doubt one of the most noteworthy cases during my tenure as police chief. Intuition, initiative, good planning, and determination were all involved, along with some top-notch crime analysis, and a healthy dose of just plain good luck. If you are a crime analyst, a detective, a street cop, or a even a deskbound chief, it just doesn't get much better than this.

Here is the contemporary Lincoln Journal Star news story by Margaret Reist from the day of the arrest. A more detailed account, including a description of the analysis that supported "Rolling the Dice," appears in Crime Mapping Case Studies: Practice and Research, 2008, edited by Spencer Chainey and Lisa Thompson.

HunchLab 2.0

A few weeks ago, we began piloting HunchLab 2.0 in Lincoln. HunchLab is a crime forecasting product from Azavea. HunchLab seeks to forecast crime by analyzing historical crime data and other information related to crime, such as weather, population density, and land use. HunchLab uses these data to predict the most likely areas for crime today, in eight hour blocks.

The output from HunchLab consists of a map of Lincoln containing a few dozen color-coded cells about the size of two square blocks. These are areas where the risk of crime is forecast to be heightened during the prediction period. These cells are approximately two blocks square. The colors represent the particular crime type for which the elevated crime risk is predicted.

HunchLab predictions are based on the analysis of Lincoln's crime data for the past five years. Where crime has occurred in the past provides a good clue as to where it will occur in the future. This is particularly true for recent crime, so crimes occurring in the immediate past are given more weight. Historical data also reveals patterns in time: crime has large peaks and valleys across the calendar and the clock.

Population density is an excellent predictor of many crime types, as is income. Densely packed low-income neighborhoods suffer more crime than sparsely populated suburban neighborhoods. Land use and zoning impacts crime. Retail businesses, such as convenience stores, restaurants, hotels and motels, grocery stores, and so forth draw people together so the chance of offenders and victims encountering one another is increased. In addition, certain kinds of businesses are related to a significantly elevated risk of crime at or around the business, such as bars and liquor outlets.

Essentially, when you gather all these kinds of data together, you can make informed predictions about where crime is most likely to occur. All of the factors that go into the prediction are based on research about the causes and correlates of crime. In fact, the commercial product emerged from research conducted at Temple and Rutgers. HunchLab goes a step further, and tests the predictions: how well did the cells identified by HunchLab’s algorithm perform in predicting where the crime actually occurred in subsequent time periods? One of the distinguishing features of HunchLab is the ongoing testing of the model, and machine learning that adjusts the predictions on the fly.

HunchLab is far better at predicting crime locations than random distribution. Although Hunchlab is predicting crime at places where most seasoned Lincoln Police Officers would recognize higher risk from their experience, there are other places that aren’t so obvious. It is also worth noting that not all Lincoln police officers are equally "seasoned."

We have helped out with the development of HunchLab 2.0 by providing data and feedback, and we are one of a handful of agencies piloting the application. For a more detailed description of HunchLab under the hood, the underpinning in criminological theory, and the research upon which it is based, go here.

Predictive analytics are getting a lot of attention in policing these days, but the same techniques can be applied to fire and EMS work. I blogged about this a few years ago, when the concept was still pretty new. Here in Lincoln, it is possible to predict fairly accurately both where we'll be sending fire engines and medic units, and when those responses will be occurring. We are using this knowledge more than ever to make informed decisions about our operations at Lincoln Fire & Rescue.

Tips for new analysts

Last year, I wrote a series of blog posts with ideas for new crime analysts. On a few occasions since then, I have needed to send links to someone to one or more of these posts. Since they were all on different days, they are a little hard to navigate to as a group. The purpose of this post is just to bring them all together in one spot, so that I can point people to the whole series easily.

Impromptu visit
Weekly slideshow
A place of your own
Better bulletins
Easy maps

I am happy to report, by the way, that my analyst friend in Grand Island is doing excellent work, has implemented most of these ideas, and has some more stuff of her own that we are envious of and trying to copy ourselves.

What might be useful

Yesterday, a member of the International Association of Crime Analysts posted a question on their forum:

"Currently, I do a report for the command staff and then several reports for individual divisions within the department that are geared toward that division only.  We are looking at continuing our command staff report, but then putting together a bi-weekly comprehensive report that includes USEFUL information.  Of course many of the things done now will be put in this report, but I would love some fresh ideas. I would like to hear your input on what kinds of reports and information you put out regularly that your department finds really useful.  So….what’s useful?  What just goes to the trash?"

I was going to respond with a few ideas, but decided a blog post would be more effective. In Lincoln, we use various methods for keeping our command staff informed, one of which is CrimeView Dashboard. Although I like this product, the purpose of this post is not to promote it. The concepts are not dependent on the software. We were trying to get this kind of information in the hands of our staff way before we adopted CrimeView. These descriptions may give you some ideas on what kinds of information might be valuable to police commanders. 

We organize these data into widgets, which reside on pages, which are within books--all electronic, and all updating automatically. The books are geographic: one for each of the department's command areas, and one that contains citywide data. Here's a description of the pages and widgets.

Current Trends Page
Widgets on the current trends in crime and dispatches. For example, this widget displays the workload trend in the past 14 days. Blue bars are calls for service,with the gray bars depicting the expected trend in the next 7 days, based on the past two years' history. The red line is the predicted trend line, which takes into account day of week variation:

Recent Crimes Page
Widgets for selected crimes within the previous 7 days. These are the kinds of crimes that are of particular interest in Lincoln, including gang-related offenses, retail business robberies, domestic violence, burglaries, thefts from autos, and this widget--drug-related crimes. The map is interactive: click on any icon for the details about that offense, or change the view from a map to a time-of-day chart, for example:

Persons of Interest
Widgets for the registered sex offenders, parolees, offenders on furlough from prison, drug court clients, gang members, and this one--registered sex offenders who have new addresses within the past week. Again the map is interactive, if you are in the app, you can click for the details::

Places
Widgets about problem places, including such things as recent crimes at schools, addresses where we have responded to party disturbances, and this widget--a graduated symbol map of places with four or more false alarms in the past 90 days. We do a lot of work aimed at reducing false alarms, and this widget shows you the problem places right away. If you were actually inside the app, instead of this screen shot, you could click on the maximize button at the top right to take the map full screen:

Wanted Persons
Widgets for people with BOLOs and arrest warrants, over various time period lengths, such as this one--arrest warrants issued in the past week. I am a big believer that keeping pressure on people with arrest warrants has a pay off. You can click one of those buttons at the top of the frame if you would prefer to see these data as a table, rather than a map:

Part 1 Crime
The current trend in Part 1 offenses, compared to the same time period last year. In this screenshot, the data is rendered as a bar graph by offense type, but in the actual app, you can click one of those buttons at the top left for a day of week chart, an interactive map, a table of the data, a time of day chart, or a temporal heat grid:

Special Interests
Widgets for several things not categorized elsewhere, such as recent stolen gun cases, liquor license violations, and this widget of assault on police officers in the past 4 weeks:

Another way we deliver information to commanders is through threshold alerts. This is a snapshot of the citywide alerts in CrimeView Dashboard. When the threshold is exceeded, the icon turns red and pulsates, indicating at a glance something out of the ordinary. Commanders are interested in emerging problems that might be lost in the sheer volume of daily activity. That's the purpose of these alerts. Only one of these (graffiti vandalism) had fired off this morning. You can click on this screenshot for a larger view that is more legible:

Volume and proportion

A question posted on the International Association of Crime Analysis forum last week caught my eye. It was by a detective at a municipal police department in Texas, who was trying to figure out how to create a map which would depict the relative number of thefts, burglaries, and robberies in each of his city's three police districts.

Since he wanted to do this using ArcGIS (something I'm pretty comfortable with), I replied to the list with my suggested method. I suspect crime analysts who lurk on the IACA list are somewhat shocked when a person in executive management responds with a solution to a technical GIS issue. I'm probably older than most of their fathers, too, but from time to time I like to remind myself that I can still do things like create pivot tables, hammer out a little html code, and wrangle a GIS project to make it do what I want it to do.

At any rate, this is similar to what he was trying to accomplish. It is a map of robberies in Lincoln from 2009 through 2013, depicted as five pies, one for each of the Lincoln Police Department's command areas. The differing size of the pies reflects the relative number of robberies within each Team, while two slices of each pie are business robberies and non-business robberies.

It's a simple graphic that at a glance conveys information about both volume and proportion. As you can see, the Southwest Team has a much larger volume of robberies, and the non-business slice is way bigger than most of the other teams. Basically, street robberies are the issue in the Southwest and Center Teams, while business robberies (though still fewer than non-business) are more prevalent in the Northwest, Northeast, and Southeast teams.

Click image to enlarge

For analysts wishing to create this effect with ArcGIS, take your polygon layer (in my case, the five police teams) and add fields for the pie slices and their values (in this example, business robberies and non-business robberies). Under "Properties" change the symbology to "Charts" and select a pie chart. A stacked bar chart would also work well with these data. There are other settings for various options, but you'll get the drift.

FireView Dashboard

This week, we have begun to roll out a new application at Lincoln Fire & Rescue, FireView Dashboard from the Omega Group. We have been using their police application, CrimeView Dashboard, at the Lincoln Police Department for several years. These products are examples of many in the marketplace for both government and the private sector that provide tools for analyzing, visualizing, and reporting data from large back-office databases. The object of such products is to provide views of data that are more intuitive and meaningful than green-striped paper spitting from a line printer.

Here are a few examples of things that I've learned or had reinforced by FireView Dashboard as I've worked on setting up data views in the past few weeks. These individual view frames are called widgets, and we have authored 391 of those for the application. That is only the pre-configured portion, though, as the Analysis Mode allows users to create a virtually unlimited number of queries on the fly. The images are just screen shots; in the app you can zoom, pan, change the view (for instance from a map to a day-of-week chart), and drill down as far as you wish--even to the individual incident data. I don't think any of these are a big surprise to firefighters in Lincoln, and some were things that I already knew or suspected, but they are now immediately clear. Click on any of these images for a slightly larger and more legible view.

Medical emergencies dominate
EMS is nearly 80% of our business, and the remainder is composed primarily of false alarms and standby duty at athletic events. Fires are few. I think of them as low frequency, but high-criticality events, for which you must be properly equipped and trained, despite their rarity. Since our core business is emergency medical services, our recruitment, hiring, training, and organizational culture need to reflect the reality of what we actually do. This ought also inform our decisions about what to drive.

The rhythm of the day
The demand for LF&R services peaks at about the same time as the demand for police services: around the time of the evening rush hour. We should keep this in mind when planning discretionary activity like a workout or fueling up the rig during this time, in order to maximize our availability and minimize turnout time.

Problematic false alarms
I've blogged on many occasions about efforts by the police to reduce chronic false burglar alarms, and the value thereof. A burglar alarm, however, is small potatoes compared to a false fire alarm, which sends a convoy of resources by comparison: six to eight firefighters in at least two large apparatus that suck diesel fuel at around 3.5 MPG. If we do something in collaboration with the owners or managers to reduce repeat false alarms, it  be a will be a very good thing. It's not just the resources, either. I walk by the photos of three firefighters and three police officers nearly every day, who were killed in the line of duty during emergency driving. Reducing unnecessary Code 3 driving protects everyone.

Residential fire hotspot
House fires have a strong geographic pattern in Lincoln, one that I have seen before. The same pattern is evident in crime. Despite the conventional wisdom, I do not believe this pattern of house fires is the result of newer construction at the edge. Rather, I think it is a reflection of population density and a host of socio-economic factors. The historic core of Lincoln, represented by an ellipse that runs roughly from Cooper Park in the southwest to the edge of Nebraska Wesleyan University in the northeast, is where we need the most fire suppression capability, and a great place to focus fire prevention work.

These are only a few examples of how we can use data and analysis to help guide our operations. FireView Dashboard is a powerful application with immense capabilities, and this is not even scratching the surface. Anyone who has read my blog for a while knows that I am an advocate for, and practitioner of, data-driven management. In this day and age it is critical for public-sector organizations to make decisions based on analysis, and to optimize their service-delivery wherever possible.

Suspended but undeterred

One of my many, many pet peeves is suspended driving. People lose their driving privileges for a number of reasons, but the leading causes are too many tickets in too short of a time period, conviction of drunk driving, and failure to have insurance after being involved in a traffic crash. Oh, and when you're convicted of driving while suspended, your driving privilege is...suspended some more. 

Yesterday morning around 5:00 o'clock I was perusing the recent police incidents, when I noticed that there had been seven suspended driving arrests since midnight. That piqued my curisosity, so I launched a query, and found that there had been 481 suspended drivers so far in 2014, and 3,153 in calendar year 2013. It has become one of the most frequent of all traffic offenses.

 Suspended drivers particularly annoy me because many are chronic repeaters, and many (probably most) do not have valid insurance, as required by law. Good luck if you get involved in a crash with one of these motorists--who by definition have already proven themselves to be poor drivers. My big concern are the chronic repeats: people who get arrested for DUS over and over again.

To my way of thinking, after you've been convicted of suspended driving a two or three times, your next conviction ought to involve the forfeiture of the car you were driving--if you own it, or if the person who owned it knew that you were suspended. Interesting pattern evident on these charts pertaining to the time of day and day of week of those 3,153 suspended driving cases in 2013. You think there might be a link to alcohol?

By the way these are screen shots from CrimeView Dashboard, which allowed me to do this analysis in a matter of moments in my living room while sipping my first cup of coffee.

Evaluation important

I only participate in one web forum, sponsored by the International Association of Crime Analysts. Once or twice a week, I will look over the posts, and from time to time I'll send a message off-list directly to someone who has posted a question. Occasionally, if I think something might have general interest to the group, I'll post my reply to the list.

Such was the case earlier this week, when an east coast analyst sought advice on how she might best contribute to a quarterly meeting at which police command staff members would be discussing proactive strategies to impact crime. This is a topic near and dear to me, that consumes a lot of space on the Director's Desk. Clicking the "Crime Analysis," "POP," or "Crime Prevention" tags in my label cloud would bring you to scores of posts on this topic.

My advice to her was to make friends with the Problem Oriented Policing Center's website, popcenter.org, and to help her commanders devise simple, straightforward evaluations to proactive strategies they implement. Evaluation is an important, yet often overlooked step in problem solving. I'm a huge believer in evidence-based practice, which demands assessing results so you can determine what works.

I have found a basic multiple time series to be one of the most practical ways to answer this question. In her post to the IACA forum, the analyst used the example of a preventative strategy to reduce larcenies from automobiles at certain apartment complexes. I'd think of evaluating that by collecting data on the volume of these crimes around the target locations before and after the intervention, then comparing that with the same time frames at other similar apartment complexes, and in the city as a whole.

In a couple of weeks, I'll be teaching my regular two-day course on this topic to recruits in the police academy, where two of the key questions they will grapple with in such scenarios are these: How will you know if it worked, and where will you get the data?

Epidemiology of police assaults

Finally, after 1,200 posts and nearly seven years, I get to use the word "epidemiology" in a title. My opportunity arises now that Drs. Joel Caplan and Phillip Marotta have published their paper, Felonious Assault and Injury to Law Enforcement: Epidemiology and Spatial Risk Factors. It is a quick read, and worthwhile for police officers to peruse.

The researchers have examined data on felonious assaults of police officers and identified spatial and temporal correlates and risk factors. I like their format, where these findings are distilled into "implications for practice"--observations and recommendations set apart in gray boxes within the report. While much of this may be common sense for seasoned officers, it's still a valuable reminder.

I made a minor contribution to the content by providing some data and very rudimentary analysis to Dr. Caplan last fall, described in this post from October 16. Lincoln's data ends up as a case study on page 13 of the report. I'm very interested in research, and really enjoy the occasional opportunity to contribute to the body of knowledge in public safety.

Beware the wee hours, officers, and do not let your guard down at facilities where you may have a certain comfort level due to your familiarity with the surroundings, such as hospitals, detox centers, the jail, and--the police station.

Not quite as intense

Last month, I wrote a post called Micro-places in which I tried my own version of some research that studied how crime concentrates in very small places: in this case, a study of individual blocks in Seattle. My little project basically looked at the same thing in Lincoln, and showed similar results: intense concentration of crime on the "top blocks."

The post was certainly not my first about the geographic dynamics of crime in Lincoln, but for some reason, it got more attention than most, as evidenced by the online comments, and also by a few delivered to me personally. One of those came from a city council member, who correctly perceived that a lot of those "hot blocks" were in retail areas.  This is really pretty obvious when you think about it.  Many common crimes are related directly to retail activity (shoplifting, credit card fraud, forged checks, etc.), and many are related to the processes that bring people together in the same place at the same time: potential victims and potential criminals--something that happens naturally around busy retail areas.

The council member wondered if the picture would be different if I had excluded those retail-oriented crimes. I opined that it would, but I have my own formula for assessing such questions, something I call the "neighborhood well-being incidents." My 2004 intern, Becky Colwell, won a national student paper competition using this concept, which she named the "quality of life index." What Becky and I did was rethink the kinds of police incidents that most accurately reflect the safety and well being in a neighborhood.

Becky's paper describes this in greater detail, but in essence we thought that some crimes had very little impact: a fraudulent credit card swipe at the grocery, a shoplifted pack of cigarettes at the convenience store, a resident of the youth detention center assaulting a staff member.  Other crimes would more obviously influence the quality of life or the well being of the neighborhood: a residential burglary down the street, a child abuse in the apartment building, a drug arrest at the park, etc.. Using combinations of incident codes and location codes in the police database we excluded the kinds of crimes at places that seemed to have little relevance to well being, which in the process also emphasized those that do.

I ran the same GIS process as in my earlier Micro-places post, but this time filtering the incident and location codes using my neighborhood well being query. The resulting map is below, and a side-by-side comparison will show some similarities and differences. The key difference, though, is not really depicted by the map so much as the data. The neighborhood well being incidents are not nearly so concentrated as crime generally. Removing the retail and institutional crime reduces the intensity of the concentration. Whereas 81% of all crime was concentrated on the top 5% of the blocks, when only neighborhood well being incidents were examined, 49% of those incidents were concentrated on the top 5% of the blocks. That's still some major concentration, nonetheless. Some of that is simply a byproduct of population density, some is related to economic and demographic conditions, and since many crime types are integrally related to our automobile culture, traffic density is another important factor in concentrating crime along certain street segments.

Micro-places

Longtime readers of the Director's Desk have been bored stiff from time to time with my ruminations about such topics as evidenced-based policing, situational crime prevention, and the like. Of particular interest to me is a phenomenon that is at the heart of a lot of good police work: digging into crime where it actually occurs--particularly impacted neighborhoods, blocks with pernicious problems, specific addresses with repeat offenses, even an individual apartment.You will find lots of examples here in my blog about problem-oriented policing strategies that are targeted at these micro-places.

A friend is the co-investigator on several research projects that have involved crime hot spots.  One of these that caught my attention last year was a study of juvenile crime in Seattle. The reserarchers, Drs. David Weisburd, Elizabeth Groff, and Nancy Morris, found that a huge percentage of juvenile crime occurs on a very small percentage of the block faces in Seattle (a block face is both side of one street, between the two adjacent intersections--like the 1500 block of S. 9th Street.)

I've been meaning to look at this in Lincoln for some time, and I got motivated yesterday to do so, a job that took about 15 minutes to set in motion, and ran on my computer while I was away at lunch.  Here's my method:  I took all the crimes reported to LPD in 2012 (21,153 total), and joined those points to the nearest street segment.  Street segments in our geographic information system are essentially the equivalent of the block faces studied by Dr. Weisburd, et al. in Seattle. I then summarized the crimes by street segment, and joined the results back into the streets layer. My result was a GIS layer of streets, where each segment also has a COUNT field that is the sum of the 2012 crimes along that segment.  Here are the results:

There were 13,771 total street segments in the City of Lincoln.  Of those, 7,481 (54%) had no crime at all in 2012. Moreover, 81% of all the crime (14,149 offenses) occurred on only 5% of the street segments (689 segments). The top 1% of the street segments (138 segments) accounted for 7,148 crimes--41% of the total.

These data are consistent with other studies, and demonstrate something very clear: crime is intensely concentrated at these very local micro-places. Police efforts focused on intervening in the conditions and preventing crime at individual premises on those blocks where crime is concentrated are far more likely to be effective than driving around aimlessly burning fossil fuel (or as we call it in policing, "routine patrol").

Here's a snapshot of the 5% of the street segments in Lincoln where 81% of the crimes occurred in 2012.

Easy maps

I wrap up the series this week of "quick wins": tips for a new crime analyst, looking for some projects that can be spun up fairly rapidly and at low cost.

Maps are useful for visualizing crime, which explains why police officers have been sticking pins in maps since the early 19th century. The human eye is a powerful instrument, able to discern information from a map that would be far more difficult to conceptualize in a list, table, or stack of paper. These days, geographic information systems (GIS) enable some very sophisticated analysis and visualization. Jessica, the new crime analyst at the Grand Island Police Department, understands this, and it was her exploration of GIS software that caused her to contact me in the first place.

Implementing GIS for crime analysis is a very worthwhile undertaking, but will require some significant time and resources.  In the meantime, though, some simple yet useful maps can be produced using free and easy tools. This has changed dramatically in the past few years. For a new analyst who has yet to acquire the software and training she will need for geographic analysis, these free services could be used in the interim to produce some simple maps of interest to officers.

Two good examples are cartodb and batchgeo. These services let users drag-and-drop tabular data that contains addresses onto a webpage, which geocodes the data and produces a credible map in a matter of seconds. The free version usually has some limitations to encourage you to spring for the pro version (such as the ads in this sample), but the free service could still have a new analyst producing a useful map or two every week.  As an example, I went to the Fargo Police Depaartment's website, screen-scraped the dispatch data from April 1st, dropped it into batchgeo, and created this map in less than five minutes of traffic crashes on April Fool's Day in Fargo. It's interactive: pan, zoom, try Pegman, click on an icon.



View April 1 Crashes in a full screen map

You can also build maps for free with Google My Maps, ArcGIS online, and others. You could make a quick map, take a screen shot, and paste it into a bulletin, or onto a PowerPoint slide to provide a visual cue to officers to go along with the text.  In Grand Island's case, they can create basic maps within their records management system, and they already have a public-facing crime map, as they participate in crimemapping.com from the Omega Group.  

Like any other Ciy whose data is published to crimemapping.com, Grand Island officers could download the CrimeView NEARme app for Android or iOS, and have a pretty useful tool even though it would be displaying the public crimemeapping.com data until and unless they signed up for the real McCoy. Those data, even though mighty basic and cleaned for public consumption, still have some value to police officers.  

The Series:

Impromptu visit
Weekly slideshow
A place of your own
Better bulletins
Easy maps

Better bulletins

This post continues this week's series on some tips for a new crime analyst.

The crime bulletin is a bread-and-butter product for the crime analyst. We all create them. You know the drill: you're the only person at the PD who actually reads all the reports every day. During the process you find interesting intelligence, uncover links between cases, develop potential suspects, and find patterns and trends before they become well known. When these findings have a broad audience of interest in the department, you commit the information you've gleaned to print, perhaps supplemented with a graph, chart, map, or photo, then copy your publication to stuff into the mail cubbies.

What I want to do, however, is to build a better bulletin by following four steps. First, KISS.  Keep it simple, seriously. Bulletins should be short. Two pages is usually one too many. Three is way too many. There may be a place for an analytic report of greater length, but a crime bulletin ought to be something you could read while standing at the urinal. Fewer words and more white space makes for a better bulletin.  If you use graphs and charts, make sure they are easily understood. This is the Achilles heel of many crime analysis products that are just too obtuse. The acid test is to hand the graph to your teenage daughter and see if she can tell you what it means, without any knowledge of the subject matter whatsoever. You'll find hundreds of graphs and charts in my previous blog posts, and most will be mighty simple.

Second, create a brand. Find a logo you like. Make a masthead, a header, a footer, a text box with disclaimers, confidence and sensitivity level, distribution restrictions, your contact information, or whatever you wish to include, but avoid unnecessary and redundant clutter.  Create a template with these elements. Use that consistently on all your bulletins. You'll find lots of examples on the website of the International Association of Crime Analysts (IACA), which, by the way, you should definitely join. For $25 per year, it's well worth it for the listserv alone, which buzzes with tips and solutions.  IACA has great publications, training of all kinds, fabulous resources for members, a certification program if you are so inclined, a fine annual conference, and more.

Third, optimize your bulletins as electronic documents.  Think of your bulletins as more than a piece of paper and supplement the printed word with interactive content.  My blog is a good example.  Most posts contain several hyperlinks that provide more information or supplement the text. Publish your bulletins in Adobe Acrobat portable document format, the ubiquitous .pdf. The Adobe Acrobat reader is free and as near universal as you can get. When relevant, build hyperlinks into your bulletins. This is easy to do in word processing or desktop publishing software, and your officers will know what to do with the blue underlined text in an online .pdf.

Posibilities for links abound.  Tag your name in the footer as a link to your email address. Tag addresses in your bulletin to Google Maps or Bing Maps to that precise address, with or without StreetView or Bird's Eye View. To do so, open the map and navigate to the spot you want. You can turn StreetView on in Google if you wish, or Bird's Eye view in Bing.  Look for the link icon in Google, and the Share button in Bing, and copy the URL. Next, highlight the address text in your bulletin, and paste the URL as the hyperlink. The text will be a blue hyperlink, just like all the links in my blog. Does your bulletin contain an out-of-state license plate number? Find a photo of that State's plate on the web, copy the URL and turn that plate number blue! While your at it, highlight the vehicle type and link to a beauty shot of the particular make and model. Here's a tip: most free .pdf distillers will not preserve the hyperlinks when you convert the .doc to .pdf. The full version of Adobe Acrobat works, but CutePDF (my favorite freebie) does not. If you use "save as .pdf" in Microsoft Office 2010, the hyperlinks will be preserved. OpenOffice also preserves hyperlinks when the document is saved as a pdf, and it is free.

Fourth, create an online repository of crime bulletins. It should be a link on that Intranet page you built yesterday, with the bulletins themselves parked on the same server. There are several benefits. You will have a chronological archive for future reference ("Wasn't there a bulletin about something similar a few months ago?").  You will save a tree or two. You will be able to email the hyperlink to the bulletin, rather than choking the pipe with a large attachment going to several dozen addresses.  You will be able to search the archive. The search mechanism can be quite simple. When you create the HTML page listing the links to the bulletins, include a title and keywords. Then, a user can just use CTRL-F (that's Edit/Find from the browser's menu) to search for the keyword.  It might look something like this:

The Series:

Impromptu visit
Weekly slideshow
A place of your own
Better bulletins
Easy maps

A place of your own

Create a place of your own, an informational website for the use of your police department. There are tons of websites that are useful for police officers. Most are public. Here in Nebraska, for example, you've got such things as the State Sex Offender Registry, the Department of Correctional Services Inmate Locator, the Nebraska State Statutes, the State employee directory, and the Department of Roads  highway cameras.

You need not guess what a blue triangular pill might be, wonder whether the VIN checks out, what an Oregon license plate looks like, whether a Fusion's taillights are round, what the name of that street is just to the west of West Lawn Elementary School who owns the house at 1432 Howard PL, or what's coming up on the training calendar.

This, of course, is barely scratching the surface. Everyone on the department has links they find useful in their browser's favorites folder. What sites do some of your officers, detectives, and sergeants find helpful? What sites has Deputy Chief Steve Smith bookmarked on his desktop in Anchorage? Ask around, and winnow the wheat from the chaf. You could gather up all those links, create a page full of them, plop it on a City web server, and point all the department PCs at that link as the browers' home page. Then, everyone would have the benefit of access to these useful sites.

It's easier than you think to build such a page, even if you can't afford Dreamweaver, and don't have time to take the class. If you gathered up the links in advance, you could ask around the faculty lounge at Grand Island High School, and identify a sharp student could make your page in an afternoon. And I guarantee you that one of those kids down the street at Central Community College taking INFO 1500 would be excited to do this, just for the benefit of being able to put this paragraph on his or her personal resume:
 

2013:  Intern, Grand Island Police Department   

Assigned to Crime Analysis Unit. Designed and implemented website for department's law enforcement Intranet. 

Keep it simple: nobody needs fancy-pants graphics or Flash. I want your page to work in any browser, and to wrap onto the small screen of a Kindle Fire or iPhone without much problem. Trust me, you can really do this. Basic HTML is a breeze, and you can learn it fast on the web at a million and one online tutorials.  You will quickly learn the secret to all computer programming is cut 'n paste.

Since your page 'o links would be composed of public web sites at the outset, security would not be an issue. It would, however, be even better if your City IT department would help you find some secure web space as the landing spot for an Intranet site. In the future, you're going to want to develop some of your own resources that will require that you control access to the content with basic security.  We're not talking about the blue prints to a nuclear reactor here, but you won't want the general public to be accessing things like the forms library you'll be creating on that site later this year, or your archive of crime bulletins coming up tomorrow.

The Series:

Impromptu visit
Weekly slideshow
A place of your own
Better bulletins
Easy maps

Weekly slideshow

This is the first in my series this week about some fairly simple and quick ideas for a new crime analysis unit or a new analyst.

Make a weekly slide show with PowerPoint, containing mugshots of a few people that your police officers ought to become familiar with. This might be a frequent flyer, a recently-released prolific offender, a dangerous character with fresh intelligence information, or a wanted subject that officers might unknowingly encounter. While it wouldn't have to be limited to mugshots, that's an easy place to start.

Loop the slideshow, and plop it onto a screen that officers will see: the assembly room where officers gather for roll call at the beginning of the shift would be ideal. The break room or the report room would be other possibilities. There might be an unused hand-me-down computer in a closet somewhere around the department that would work satisfactorily. If not, a low-end desktop with a  24" monitor would suffice for well under a grand.

I'd really like you to get an even bigger monitor to mount on the wall, and a good wireless keyboard and mouse, because you will eventually want to use this equipment for another purpose. There are four 60" monitors in Lincoln's assembly room, and I saw a wall-full at the Los Angeles Police Department, but in Grand Island one would do the job. I just bought a 52 inch LED for the family room for $600 at Sam's Club, and it looked great streaming the family photos at Easter. It would work just fine in Grand Island. In a city of 50,000, I imagine that somewhere between 3 and 8 officers are typically coming on duty at shift change, depending on the time and day. A set up similar to the Council Bluffs Iowa Police Department's assembly room would be perfect for such a group.

Your slide show can just run continuously when the computer and monitor are not being usedfor other  purposes. Keep the number of slides low, and resist the temptation to pack it with more information. Sometimes less is more. I'd recommend four to six mugshots--no more. Use a dark slide background, light-colored text, a subtle transition, a big font, and very few words:

Luke N. Long
Window Peeker.
Blue F-150,  BENCNU


The Series:

Impromptu visit
Weekly slideshow
A place of your own
Better bulletins
Easy maps

Impromptu visit

Last Thursday I was 125 miles west down Interstate 80 in Kearney, Nebraska to give two presentations at break-out sessions during the Nebraska GIS Symposium. As I was preparing to head home, I checked my email. Among my messages was one from Jessica, a new crime analyst at the Grand Island Police Department. She was introducing herself, and seeking some information. Grand Island was just 35 miles away, and right on my way back to Lincoln, so rather than a phone call, I just invited myself to drop by her office. I sent her a 45 minute Glympse, so she'd know when I was getting close.

Jessica is new to crime analysis, but has lots of law enforcement experience.  She wanted some information from me about geographic crime analysis software, which I was happy to provide.  More importantly, though, I gave her some advice that I have shared  over the years with many other crime analysts in a similar position: starting out in a new position in a department that previously has not had an analyst.

Even though she's been on the job for less than a month, she has a lot of irons in the fire. She's evaluating software, learning the department's records management system, she has training scheduled, and she has a lot of good ideas already. She has some very strong assets  strong education and experience, a competent records management system, a mobile network with computers and broadband in the patrol fleet, and a supportive chief.

I recommended that she focus on a few "quick wins": projects that she could pull off with very little expense, rather swiftly, that would noticeably improve the information flow within her department.  Some early success sets the stage for bigger projects that will take more time and resources. It also makes a new analyst feel good about his or her job, and starts to show officers the value of their new crime analyst!

By the time I had to bug out, I'm sure her head was spinning.  Although she was taking notes furiously, I thought it might be helpful to both her and to other new crime analysts for me to commit my ideas for a few quick wins to writing. I will be using my blog this week to do so, describing a few projects that a new analyst just starting out could get rolling right out of the gate.

There's nothing novel here, all these things are being done in lots of places around the country. Some of my colleagues will look at these ideas and surmise, correctly, that there isn't really anything analytic about what I am suggesting. This is quite true. Spatial statistics, kernel densities, standard deviation ellipses, predictive forecasts, risk terrain modeling, and such are down the road. Let's just think about what a new analyst in a city of 50,000 could get cracking on as a starting point to get more information about crimes, criminals, trends and patterns to the troops.

The Series:

Impromptu visit
Weekly slideshow
A place of your own
Better bulletins
Easy maps